Package elki.clustering.hierarchical

Class MedoidLinkage<O>

java.lang.Object

elki.clustering.hierarchical.MedoidLinkage<O>

Type Parameters:

O - Object type

All Implemented Interfaces:

Algorithm, HierarchicalClusteringAlgorithm

@Reference(authors="D. Herr, Q. Han, S. Lohmann, T. Ertl",title="Visual clutter reduction through hierarchy-based projection of high-dimensional labeled data",booktitle="Graphics Interface Conference",url="https://doi.org/10.20380/GI2016.14",bibkey="DBLP:conf/graphicsinterface/HerrHLE16") @Reference(authors="S. Miyamoto, Y. Kaizu, Y. Endo",title="Hierarchical and non-hierarchical medoid clustering using asymmetric similarity measures",booktitle="Soft Computing and Intelligent Systems (SCIS) and Int. Symp. Advanced Intelligent Systems (ISIS)",url="https://doi.org/10.1109/SCIS-ISIS.2016.0091",bibkey="DBLP:conf/scisisis/MiyamotoKE16")
public class MedoidLinkage<O>
extends java.lang.Object
implements HierarchicalClusteringAlgorithm

Medoid linkage uses the distance of medoids as criterion. The implementation is a simplified version of AGNES, as we do not have to update a distance matrix. This approach was invented at least twice. An approach more consistent with the ideas of optimum medoids is found in HACAM, which uses the loss of the resulting medoid after merging instead.

References:

D. Herr, Q. Han, S. Lohmann, T. Ertl
Visual clutter reduction through hierarchy-based projection of high-dimensional labeled data
Graphics Interface Conference

S. Miyamoto, Y. Kaizu, Y. Endo
Hierarchical and non-hierarchical medoid clustering using asymmetric similarity measures
Soft Computing and Intelligent Systems (SCIS) and Int. Symp. Advanced Intelligent Systems (ISIS)

Since:

0.8.0

Author:

Erich Schubert

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	MedoidLinkage.Instance	Main worker instance of AGNES.

Nested classes/interfaces inherited from interface elki.Algorithm

Algorithm.Utils

Field Summary

Fields

Modifier and Type	Field	Description
protected Distance<? super O>	distance	The distance function to use.
private static Logging	LOG	Class logger

Constructor Summary

Constructors

Constructor	Description
MedoidLinkage(Distance<? super O> distance)	Constructor.

Method Summary

Modifier and Type	Method	Description
TypeInformation[]	getInputTypeRestriction()	Get the input type restriction used for negotiating the data query.
ClusterPrototypeMergeHistory	run(Relation<O> relation)	Run the algorithm

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface elki.clustering.hierarchical.HierarchicalClusteringAlgorithm

autorun

Field Detail

LOG

private static final Logging LOG

Class logger

distance

protected Distance<? super O> distance

The distance function to use.

Constructor Detail

MedoidLinkage

public MedoidLinkage(Distance<? super O> distance)

Constructor.

Parameters:

distance - Distance function to use

Method Detail

run

public ClusterPrototypeMergeHistory run(Relation<O> relation)

Run the algorithm

Parameters:

relation - Relation

Returns:

Clustering hierarchy

getInputTypeRestriction

public TypeInformation[] getInputTypeRestriction()

Description copied from interface: Algorithm

Get the input type restriction used for negotiating the data query.

Specified by:

getInputTypeRestriction in interface Algorithm

Returns:

Type restriction